Note: this e-mail is on a public mailing list. On Thu, Sep 3, 2009 at 5:42 PM, Luis R. Rodriguez<mcgrof@xxxxxxxxx> wrote: > I'll soon commit what I have so far, as soon as it compiles. Then general > help with mac80211 stuff will be greatly welcomed. OK I just committed this stuff as a new file as ath9k_htc.c. I'm not sure if that's the best name given that htc.c may need to be renamed later to ath_htc.c in case of name collisions with any other htc.ko stuff in the kernel. I am not aware of any htc.ko though so perhaps we're in luck. Anyway -- ath9k_htc.c should give you an idea of how things will work -- HTC will provide the host <--> target communication while direct hw acess will be issued through code already in place on ath9k in the files: drivers/net/wireless/ath/ath9k/btcoex.c drivers/net/wireless/ath/ath9k/calib.c drivers/net/wireless/ath/ath9k/eeprom_4k.c drivers/net/wireless/ath/ath9k/eeprom_9287.c drivers/net/wireless/ath/ath9k/eeprom.c drivers/net/wireless/ath/ath9k/eeprom_def.c drivers/net/wireless/ath/ath9k/hw.c drivers/net/wireless/ath/ath9k/mac.c drivers/net/wireless/ath/ath9k/phy.c this includes ath9k_hw_* access stuff. This is what we typically would call at Atheros the 'HAL' but since we no longer have a 'Hardware Abstraction Layer' I will refrain from using that further here. What will differ between ath9k and ath9k_htc will be the harware access for read/write. On ath9k we do not sleep on reads/writes, for ath9k_htc we do need to sleep. We'll have to figure out a way to allow for each driver to have their own read/write op. Technically we could have just used ath9k for the mac80211 part but we would have then probably have had to re-write a lot of the core code which currently does not depend on sleeping during read/write. So -- we will be building the above C files together with ath9k_htc.c -- eventually we will share this upstream with ath9k. We can do this either with a direct commit of the above files locally onto the ar9271 git tree or use a script to update for us based on our local tree. I'm inclined to just commit th files locally for now, modify the headers when needed and send patches as we to accommodate ath9k as we go. Let me know what you think. Luis _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
